Welcome to the Q&A series.
Today let’s welcome Thomas Leitner. Thomas is the head of the computer group of the mathematics department of the University of Vienna (in Austria, Central Europe) and better known as the author of the kramdown library and tool - a fast, markdown-superset processor (in ruby) for converting plain old vanilla text into hypertext (powering GitHub Pages, Jekyll, and friends).
Q: Can you tell us a little bit about yourself? How did you get started with Ruby?
A: Ever since my parents got a computer when I was in middle school I was fascinated by it. As this was pre-Internet (at least at the beginning) I learned by watching my older brother use it and got the hang of it quite fast.
The most interesting thing was programming which developed into one of my favorite hobbies. First I used Basic and Pascal, later Delphi and in 2001 while studying Software Engineering I found Ruby and quickly fell in love with it.
Though not perfect, Ruby is nearly perfect, at least for me, and all the programming I do in my spare time is in Ruby nowadays.
Q: Do you remember - how did you find out / get started with markdown? Did you try out any alternatives e.g. textile, wiki text, etc.?
A: I was creating a static website generator, webgen, back in 2003 and didn’t want to write the content in HTML. At first I used plain HTML but later I discovered Textile and used it as main way for writing the content.
However, I discovered Markdown and the Maruku library soon afterwards and it seemed a much better language for the purpose than Textile. So Textile was demoted and Markdown took the place as default markup language.
Q: How (and why) did you get started / end up writing your own markdown converter, that is, kramdown?
A: Maruku, the Markdown conversion library written in pure Ruby that I used for webgen, wasn’t maintained anymore and there still were many bugs to be fixed.
I looked at the source of Maruku but soon decided that I didn’t want to maintain it because it wasn’t easy to update or extend.
However, I really liked the additional features that Maruku brought to the Markdown world (like inline attribute lists) and therefore did what every hacker would do: Implement his/her own solution since no other Ruby library had the same feature set.
So I started to lay out a simple, non-formal specification of what kramdown should do for various elements and began implementing it. Once I had the most important elements and the basic thing running, I put the code on Github and released it.
Q: Can you tell us some challenges you faced writing kramdown?
A: Since Markdown doesn’t have a real, formal specification, I had to pick and choose features and edge case implementations from the available libraries. And it wasn’t always easy to decide what to use and what not to use, and sometimes I implemented a certain feature in a completely different way when it seemed the better solution.
Another challenge is that when you program in Ruby it is easy to implement things in an ingenious way, like using meta programming and such. However, after you program a long time in Ruby you realize that most of the time you are better off avoiding these solutions.
Q: Do you have any favorite markdown extensions e.g. todo lists, meta data, emojis, auto links, etc. - or worst?
A: My favorite extension to the original Markdown are the inline attribute lists. These allow you to specify any attribute on any element, with special shortcuts for classes and IDs.
This was also the reason why I really wanted to use Maruku instead of other libraries such as BlueCloth, and the main reason why kramdown came to be.
Q: Any thoughts on adding / getting multi-line tables working in kramdown - works great in wiki text e.g. wikipedia or something ;-) ?
A: Maybe in version 2.0 but one of the goals of Markdown is that it should be easy to read and write. So if a syntax doesn’t feel to be Markdown-y I probably won’t include it.
Q: What’s next for kramdown? Any plans for 2.0?
A: I consider kramdown a mature and stable library. If you read the release announcements you will find that there have been regular improvements and bug fixes but nothing really major. And I think that is good for kramdown’s users because they can depend on it and the versioning scheme.
As for 2.0: Since this is a major release there will be breaking changes, not just in the internal structure (I will probably remove some of the meta programming stuff) but also in the syntax.
One thing that is sure to change is the current table syntax. Support for strikethrough and superscripts/subscripts will also probably be incorporated.
However, there is no date set for when 2.0 will be released or even worked on. This depends on the amount of spare time I can dedicate to the development of kramdown.
</>
Thanks for your time and answers and - of course - thanks for the great kramdown library and tools!
Links / References:
kramdown & webgen
Markdown
Textile
Miscellaneous